There is a moment in nearly every CB550 cafe build where the stock four-into-four comes off. It's heavy, the cans are probably weeping rust at the seams, and it is the single biggest visual obstacle between the bike you have and the bike in your head. A four-into-one goes on in its place. The 550 loses real weight, gains a voice, and you push it out of the garage feeling like you've finally done something.
Then, about four miles in, you find the flat spot.
What happens next has been documented on the SOHC/4 Owners Club forums more thoroughly than anywhere else, in a ten-page, 249-reply, fifty-three-thousand-view thread that started in August 2013 with one man, a '76 CB550K and a brand-new stainless MotoGPWerks four-into-one. It's the closest thing the small-block world has to a controlled experiment on a single question: what does a CB550 actually need when you change only the exhaust?
And running alongside it, mostly unnoticed by the people asking that question, is a forty-page thread with nearly a thousand replies about what happens to the pipe you just took off. Read together they make an argument that no single post in either thread quite makes.
The argument at the centre of it
The thread's opening question was the ordinary one — what jets do I need? — and within two days it had produced a genuine disagreement between two of the board's most-read technical voices. It's worth laying out properly, because it's the argument you'll have with yourself in the garage.
The first position is the clean one, and it has the appeal of simple physics. If you bolt on an aftermarket exhaust and leave the intake completely stock, no jetting changes are needed. No extra air is coming in, so no extra fuel needs to go with it. At most you give the mixture screws a quarter turn richer. Change the intake — pods, stacks, a hacked airbox — and the whole game changes. But a pipe on its own doesn't move fuel requirements.
The rebuttal came from one of the forum's long-standing tech regulars, and it didn't argue theory. It argued Honda's own parts numbers, and it's the most useful thing in the thread.
Consider the two CB550s Honda sold in the same years. The 74–76 CB550K left the factory with a four-into-four exhaust, a #100 main jet, and the needle clipped in the fourth position from the top. The 75–77 CB550F left the factory with a four-into-one, a #98 main jet, and the needle in the second position from the top. Same engine. Same cam. Same induction system ahead of the carbs. The only meaningful difference is the exhaust — and Honda jetted the four-into-one bike leaner, not richer.
The reasoning offered was that the F's quieter, higher-pressure muffler scavenges the cylinder slightly less completely. With a little more of the previous burn's leftovers still in the chamber, the next cycle wants more oxygen and less fuel. You trade a bit of peak power near redline for better efficiency in the midrange, which is where street bikes actually live.
The same voice offered a second data point that's hard to wave away. The 77–78 K models kept a four-into-four but with reduced-diameter outlets — higher exhaust pressure again — moved to PD carbs, and came with #90 mains plus a proper idle circuit. Nothing ahead of the carbs changed. Honda raised exhaust pressure and dropped fuel, deliberately, chasing the lean burn the EPA was asking for.
So the honest conclusion isn't "a pipe always needs bigger jets." It's something more useful and considerably more annoying: changing exhaust pressure does change what the engine wants, and the direction depends entirely on the pipe. A free-flowing aftermarket four-into-one lowers back pressure, which pushes the opposite way from Honda's F-model logic — and nobody selling you the pipe has measured what it does. That engineering detail, as the thread put it, gets handed to the buyer.
What people actually ended up running
Which is why the most valuable content in ten pages isn't the theory. It's the accumulated settings from 550 owners who did the work, and they cluster tightly enough to be genuinely useful as a starting point.
Stock for the 022A carbs on a 550 is a #100 main, a #40 or #38 slow jet, mixture screws 1.5 turns out, and the needle in the fourth clip position.
The original poster finished — after months — at a #105 main, #38 slow jet, 1.25 turns out, and the needle moved up to the third position from the top, running a Uni foam element inside the stock airbox. A member in Ohio at 400–800 feet with the same pipe on a '75 550 and the stock filter landed at #105 mains, #40 idles, 1.25 turns, needle in the middle, and reported no hesitation at all once warm. A third member with the identical setup to the OP found that the stock fourth-notch needle position ran rich and sooty at small throttle openings; moving to the third notch cleaned it up completely, though he then suspected he'd gone slightly lean up top and was planning to shim the needle half a notch to split the difference. A 550 running the same pipe with pods — a different animal, and jetted accordingly — settled at a #110 main, #38 pilot, one turn out and the needle in the third notch.
Two details from that spread are worth writing on the airbox lid. First, the difference between a #38 and a #40 slow jet was not subtle: the OP found the 40s gave poor running and black deposits while the 38s were smooth with only faint popping. On a 550 with a free pipe, one jet size on the pilot circuit is a real change. Second, almost everyone who got a good result moved the needle, and several got more improvement from the needle than from the mains.
Which points at the thread's most repeated correction: the main jet mostly controls wide-open throttle, and you are almost never at wide-open throttle. Something like eighty percent of real riding happens at idle and in the midrange — even on the freeway you're nowhere near WOT. If your complaint is a stumble pulling away from a light, the main jet is the wrong place to be looking, and going up two sizes on the mains to fix a low-speed hesitation is how people end up chasing their own tail for a season. The OP did exactly that: went from #100 to #105 and immediately reported worse hesitation down low, then spent the next page confused about why a main jet change had affected the bottom end at all.
The plot twist: it wasn't the jets
Here is the part that makes this thread worth reading all the way to the end, and it's tucked into an edit at the very top of the first post, dated four months after the thread started.
The bike was, by then, tuned very close to perfect. But it retained one stubborn fault: a small blubber or flutter at around three-quarters throttle, somewhere between six and seven thousand rpm. Many different jetting combinations were tried. None of them cured it.
What finally cured it was checking and resetting the float height to 22mm — the stock CB550 figure. The flutter vanished. The bike would take two people to eighty on the freeway without complaint.
Months of jet swapping, and the answer was a fuel level that had drifted out of spec. This is the lesson the SOHC/4 veterans keep returning to in every subject we've covered in this series, and it's never a fashionable one: a carburetor cannot be jetted around a fuel level that's wrong. The jets meter fuel relative to a height in the bowl. Move the height and every circuit lies to you at once — which is precisely what a symptom that won't respond to any jet change is telling you.
Before you buy a jet kit for your new pipe, set the floats, sync the carbs properly on a gauge rather than by bench-setting alone, and make sure the ignition side is honest. Then, and only then, start reading plugs.
The discipline the thread keeps preaching
Scattered across ten pages is a set of habits that separates the people who sorted their bikes in a weekend from the people who were still posting eighteen months later.
Change one thing at a time, and change nothing else until the jetting is settled. The OP eventually fitted a set of Dyna coils while mid-diagnosis, and immediately lost the thread of what was causing what. The reply he got is the right one: resist the urge to make setup changes until the jetting is correct, and after that, one at a time.
State your altitude when you ask or answer. A member's numbers from 500 feet in Ohio are not the same numbers you want in Denver, and the thread's most careful contributor kept asking people to include it before adding their settings to the running list.
Buy an assortment, not a guess. Several members pointed to jet kits that come as a small box of sizes, and noted you can usually ask the seller to swap the assortment for the range you actually need. Ordering one size, waiting a week, riding for an afternoon and ordering another size is the slowest possible way to do this.
And read the plugs after honest riding, not after a warm-up lap. Dark, sooty plugs after a short ride tell you less than you think.
The unglamorous install details that cause most of the leaks
Buried in the first two pages is the practical stuff nobody mentions until you've already made the mistake.
You need new copper crush gaskets. They aren't included, they're a few dollars, and they're what actually seals the header to the head. Just as important: make absolutely certain the old crush gaskets came out with the old pipes. They go sooty and black and sit flush in the port, and they're easy to mistake for part of the head. A flat screwdriver felt around the flange face will usually pop them free. Fitting new gaskets on top of old ones is a miserable way to start.
Whether you can reuse them if you take the header off again is one of the thread's smaller disagreements. The pragmatic answer was that you'll usually get away with it once and should replace them if the pipes have been off several times. The stricter answer — any time you can see them, replace them — is cheap insurance for a part that costs less than a tank of fuel.
The spring-mounted, quick-release header design exists specifically so you can pull the pipe for an oil change without disturbing the collars and their gaskets. Pull the springs, slide the header off the collars, leave the seal alone.
Two other things came up repeatedly. Assemble the header complete — sleeves, collars, header — and offer the whole thing up as one unit rather than bolting the collars to the head first; cylinders one and four sit at slight outward angles and you'll be prying at the outer tubes otherwise. And a thin bead of copper RTV where the inserts meet the header is worth doing on assembly: one member had no leak at all initially and then developed one at the number two joint at around a thousand miles.
So what does it actually give you?
The most refreshing post in the whole thread is the one that answers this honestly.
Horsepower: probably nothing you can measure, and the racers would laugh at anyone who claimed otherwise. What is real is that back pressure drops enough that the engine revs noticeably more freely — the same member found himself over-revving by several hundred rpm when blipping for downshifts for the first few rides, because the throttle response had changed. There's some ground clearance. And there's weight, which on a small block is a genuine, meaningful saving.
His closing observation is the sharpest line in ten pages: the noise biases everyone into believing it's faster. Which is fine — that's a legitimate reason to buy a pipe, and better than pretending you did it for the dyno sheet. Just know which one you're buying.
The other half of the story: the pipe you just took off
Now for the thread that the people in the jetting thread should probably read next.
It opened in October 2013 with a simple question about original Honda four-into-four exhaust, ran to nearly a thousand replies over forty pages, and is still being added to a decade later. It is anchored in the CB750 world — the legendary HM300 and HM341 sets — but every economic lesson in it now applies squarely to the CB500 and CB550, and 550 owners should understand what's coming.
The pipes are discontinued and the forum's settled consensus is that Honda will never make them again. The tooling cost against the size of the market makes it a bad business decision, and as one member put it with brutal clarity: half a dozen enthusiasts on an owners' club website is not a market. No demand, no supply.
The prices tell the story. Replica "no-no" pipes — reproductions carrying no part number and no emissions paragraph — have run around two thousand dollars a set delivered, with the maker at one point advertising a batch as the last sets they would ever produce. A single near-new original number-two muffler sold on eBay for six hundred and eighty-five dollars. One member recalled buying two complete sets of genuine Honda pipes, delivered to Texas, for seventeen hundred and sixty dollars total, six or seven years before the thread started — and wished, obviously, that he'd bought more.
And here's the part that ought to make every cafe builder pause. The reason these pipes are gone is not that they wore out. It's that they were thrown away, deliberately, by people doing exactly what we do.
Dealers pulled brand-new stock exhausts off brand-new bikes and fitted four-into-twos and four-into-ones, because that's what customers wanted and that's what looked right. One member admitted doing half a dozen sets himself at the shop where he worked. A shop foreman at one dealership took two pickup loads of takeoff pipes to the scrapyard — off racks that covered two walls. Another member found three brand-new pipes hanging in the rafters of a scrap steel building; the fourth had been used as a pry bar.
Everyone in 1975 was completely certain those pipes were worthless.
There is also a practical note in there worth acting on today. The rust story is not uniform: the header sections rarely rot from the inside and are easy to keep in good order, while the mufflers are what die — and short rides are the killer, because the cans never get hot enough to drive the moisture out. One member sold the header portion of his original system after seventy-five thousand kilometres because it was still effectively as new, even though the cans were finished. So a set of stock pipes with dead mufflers is not necessarily a set of scrap.
The through-line
Put the two threads side by side and the advice writes itself.
Fit the four-into-one if you want it. It looks right, it sounds right, it takes real weight off a small block, and the forum's accumulated 550 settings give you a starting point that didn't exist twenty years ago: expect to end up somewhere around a #105 main and a #38 slow jet with the needle raised a notch if you're running the stock airbox, and expect to arrive there by moving the needle at least as much as the mains. But set your float height and sync the carbs first, because otherwise you'll spend a season buying jets to solve a problem that lives in the bowl.
And do not throw the old exhaust away. Do not sell it for scrap weight, do not leave it behind the shed for three winters, and do not let the mufflers decide the fate of the headers. Wire-brush it, oil it, box it, label it, and put it in the loft. The stock CB500 and CB550 four-into-four is on the same road the 750 pipes went down — discontinued, unreproduced, and steadily disappearing into skips during exactly the cafe racer boom we're all still part of.
You are not obliged to keep your 550 original. But somebody, thirty years from now, will be very glad that you kept the parts.
Credit where it's due
None of this is ours. It belongs to the members of the SOHC/4 Owners Club community, who have spent two decades swapping jets, reading plugs, arguing about exhaust pressure in public and quietly recording what worked so the rest of us don't have to buy the same education twice. Both of those threads are worth reading in full, and the forum runs on donations from the people who get value out of it. There is no substitute for an archive that deep.
When the pipe is on, the jetting is sorted and you're looking at what else the bike needs, that's where we come in. The Thirteen Motorcycles CB550 catalog is built for exactly this stage of the project — parts designed for this bike and made to work together, so the changes you make add up to a motorcycle that runs as well as it looks.
0 comments